1. The case study documents Gabriel P.'s weight fluctuations over 7 years where he gained 97 lbs (44 kg) in 5 years due to underestimating his true metabolic rate by 200 calories per day based on an equation.
2. Gabriel then lost 88 lbs (40 kg) in 2 years by accurately measuring his resting energy expenditure of 1900 calories per day and creating a 500 calorie daily deficit through an energy balance approach.
3. A clinical study of overweight and obese participants found 42% had slower true metabolic rates than estimated by equations, with differences of up to 520 calories per day. Accurately measuring metabolic rate is important for effective weight management.
2. Case #1: Gabriel P.’s case
+ 2 years
(- 40 kg)
June 2015
- 88 lbs
+5 years
+ 97 lbs(80 kg)
176 lbs
(124 kg)
273 lbs
(+ 44 kg)
(84 kg)
185 lbs
1. Why did Gabriel gain 97 lbs (44 kg)?
2. How did Gabriel lose 88 lbs (40 kg)?
2www.breezing.co
3. Mifflin - St Jeor equation: Man:
REE(M-StJ) = [10 * weight (kg)] + [6.25 * height (cm)] - [5 * age (y)] + 5
Why did Gabriel gain 97 lbs (44 kg) in 5 years?
• He used a calorie
calculator to estimate
Total Burn: 2100 kCal/day
3www.breezing.co
4. Estimated Total Burn: 2100 kcal/day
First True Total Burn: 1900 kcal/day
Difference between Estimated - True Burn:
200 kCal/day
How does this difference translate to weight?
[(200(kCal/day)*7*52weeks/year)]/[3500kCal/lbs]=
+ 20 lbs/year
5 yrs. ~100 lbs Total ~ 45 kg
Why did Gabriel gain 97 lbs (44 kg) in 5 years?
+5 years
+ 97 lbs
80 kg
176 lbs
124 kg
273 lbs
+ 44 kg
Measuring Energy Expenditure was a
key point in explaining why Gabriel
gained weight
4www.breezing.co
5. How did Gabriel lose 88 lbs (40 kg)?*
+ 2 years
- 40 kg
- 88 lbs
124 kg
273 lbs
84 kg
185 lbs
1400 kcal/day - 1900 kcal/day- 500 kcal/day ~
Gabriel expected a deficit of 3500 kcal per week
equivalent to a loss of 1 lb per week (52 lbs/year).
Gabriel’s actual weight loss was 44lbs/year, a total of 88 lbs in 2 years
Energy Balance Equation
Energy Storage = Energy Intake - Total Energy Expenditure
Initial approach
5
*Dr. Pablo Pelegri (MD), Dr. Liliana Balsells (MD), Buenos Aires, Argentina; Breezing’s user experience team.
6. 1400 kcal/day - 1900 kcal/day- 500 kcal/day ~
Energy Balance Equation Components
Energy Storage = Energy Intake - Total Energy Expenditure
200 kcal/day1700 kcal/day
=
Resting Energy Expenditure represents a large percentage (75-95%) of Total Energy Expenditure
- [ + ]
Resting Activity
Knowing Resting Energy
Expenditure was a key point
for Gabriel
(89%)
6
7. How many cases like Gabriel’s are out there?
124 kg
273 lbs
7www.breezing.co
8. Characteristics of the population
Dr. Craig Stump, MD
8
www.breezing.co
Case #2: Clinical study in an overweight and obese
population*
* Most of participants had T2 Diabetes, or were at risk of Diabetes
9. Difference of Calculated REE* – True (measured) REE
-800
-600
-400
-200
0
200
400
600
800
1000
1200
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45
Calculated REE - Measured REE
Female
Male
DifferentialRestingEnergyExpenditure(kCal/Day)
Study Participant Number
Group A
Group B
Group C
Dr. Craig Stump, MD
Group C
42%
* Predictive Equation (Harris-Benedict)
42% of the cases in the pilot study
group (overweight and T2 diabetes)
had slower metabolic rates than
what the equation predicted
9
10. Why we can’t use equations to calculate REE ?
10
An actual REE value (from indirect
calorimetry measurement) can differ
from an estimated REE value (from
the Harris-Benedict calculation).
The results show that for people of
same gender and weight (e.g. men
and 63 kg) the difference in actual REE
values can be as high as 520 kCal/day.
If, for instance, subject A’s goal is to
maintain weight, and the estimated
REE (1640 kcal/day) is higher than the
body’s actual REE (1480 kcal/day), a
calorie recommendation based on the
REE estimate will lead to weight gain.
Therefore, accurately measuring REE
is crucial in establishing an effective
weight management plan.
Plot from J. Arthur Harris and Francis G. Benedict, A Biometric Study of Human Basal Metabolism, Proc Natl Acad Sci U S A. 1918 December; 4(12): 370–373.
Criscione, L. & Durr-Gross, M. Eating healthy and dying obese. Vitasanas GmbH, http://www.vitasanas.ch, ISBN: 978-3-0033-02225-6 (2010).
2490
2290
2090
1890
1690
1490
1290
1090
35 45 55 65 75 85 95 105
2000 kCal/day
1640 kCal/day
1480 kCal/day
520kCal/day
64 kg
REE(kCal/day)
Weight (kg)
A
Data from seminal Harris-Benedict’s work
12. ControlGroup
InterventionGroup
12
Case #2: Clinical study in an overweight and obese
population – Six-month study design
•The participants from the control
group had an iPad with My Fitness
Pal App to track calorie intake, an
activity tracker to track steps and
floors, and a weight scale.
• Each participant in the control
group was recommended a 500-
calorie deficit intake based on the
Harris Benedict Equation
• The intervention group had
the same gadgets as the
control group, as well as a
Breezing Tracker.
• Both groups were followed
up with a Standard-of-Care
procedure for 6 months, and
were reached by e-mail
every 2-3 weeks with
general health information.
13. Case #2: Weight & Body Mass changes
Observation: Weight change is accounted from 1st day the participant use MFP (baseline period) up to 6 months
after the study
-50 -40 -30 -20 -10 0 10 20 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Series2
Series1
Weight change* (lbs)
Participants
Control Group (CG)
Intervention Group (IG)
Other results:
Weight loss Greater Than 6 lbs:
CG: 40% (8/20) vs IG: 68% (13/19)
Intervention Group: 17 of 19 participants (89%)
lost weight, 1 stayed steady and 1 (5%)
gained 1.9 lbs.
Control Group: 11 of 20 participants (55%) lost
weight, 1 stayed steady and 8 (40%) gained
2+ lbs.
13www.breezing.co
14. -250
-200
-150
-100
-50
0
1 2
Case #2: Weight & Body Mass Index (BMI) changesWeightchangeaverage(lb)
Group
Control Intervention
*Statistical significant (p= 0.03)
The Intervention group’s total weight
loss was 3 times greater than the
control group
The difference in BMI changes in the
intervention group was significantly
different with respect to the control
group
The intervention group’s drop of BMI
from 35.5 resulted in a change from
Obese Class II Group to Obese Class I
Group
Control Intervention
BMI:-1.9
BMI:-0.5
14
The control group’s drop of BMI
from 36.9 was not large enough
to move out of Obese Class II
Group
15. Control Group
Intervention Group
Percentage of calorie intake completed days (%)
Participants
15www.breezing.co
0 20 40 60 80 100
1
3
5
7
9
11
13
15
17
19
Case #2: Calorie Intake Completed Days*
* Completed days represent calorie intake values with equal or 25%+ of recommended calorie intake
17. Case #2: Calorie Intake Entries
0 25 50 75 100 125 150
SE
ND
CM
BA
GP
ND
JM
JG
DS
AD
LJ
JS
JJ
GV
RD
GC
FV
WN
Avg MFP
SD
Avg MFP+B
CB
DT
VV
LP
YS
SG
AR
DL
SB
AA
OF
JF
AM
MH
MS
BB
JH
JS
JG
Total Measures
Number of Entries
Participants
ControlGroupInterventionGroup
My Fitness Pal (MFP)’s Volume Entries
(including diet, activity, weight, comments)
Breezing Entries
63 = MFP’s entry average
79 = MFP’s entry average
Intervention Group:
25% more entries
than control group
www.breezing.co
18. Case #2: Benefits of weight loss in blood parameters
18
Intervention group had a better outcome
for HDL cholesterol (increased HDL
cholesterol with a significant difference of
p = 0.037 with respect to the control group
-25 -15 -5 5 15 25
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Series2
Series1Controls
Intervention
HDL change
Diastolic Blood Pressure Intervention group had a better outcome
for reduction of diastolic blood pressure: a
decrease with a significant difference of p =
0.07 with respect to the control group
19. Summary of facts from the study (Case #2)
1. Breezing users had:
i) Effectively lost more weight (89% vs 55% in the control group)
ii) Completed 70% more calorie intake enties in the calorie counting app
iii) More comprehensive use of calorie counter app via entry volumes of
diet, activity, weight, and comments.
iv) Better HDL cholesterol and Diastolic Blood Pressure parameter
outcomes
2. How does knowing Correct Calories Burned relate to Weight Loss?
89% efficiency of weight loss (IG) vs. 55% efficiency of weight loss (CG)
5% of weight gain (IG) vs. 40% of weight gain (CG)
19www.breezing.co
20. HbA1c reduction
20
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1
1
3
5
7
9
11
13
15
17
19
Series2
Series1
Controls Intervention
Controls
Intervention
Case#2 (cont.): General weight loss effect in T2 diabetes
The weight reduction resulted in a reduction of
glycated hemoglobin in both groups (p < 0.1)
Since both groups had a relatively high rate of
weight loss (89%-IG and 55%-CG), there was not
significant difference between groups in regard
to improvements of glycated hemoglobin (both
groups did improved the T2 diabetes parameter)
CONCLUSION: weight loss has an intervention
effect on lessening T2 diabetes symptoms and
decreasing the risk of developing diabetes
Between groups: no difference
21. What about pregnancy?
To learn more watch:
https://www.youtube.com/watch?v=tHS-pegE_gQ
21www.breezing.co
22. 0 20 25 30 35 40
900
1200
1500
1800
2100
2400
After birth
1540
1890 (+/-150) 1680
(+/- 50)
REE(RMR)(kcal/day)
Pregnancy week
1830
(+/- 30)
Baseline REE = 1,200 kCal/day
Cold/
Flu
April 8th, 2015
Study case #3: Resting Energy Expenditure
during pregnancy*
Jan 8th, 2015
day
How does the
profile connect
to other body
parameters?*Dr. Corrie Whisner, American Society of Nutrition's Public Information Committee
* D. Jackemeyer, BSW, Application Scientist, Arizona State University
22
23. Comparison of REE
with Weight
Mifflin - St Jeor equation:
Woman:
REE(M-StJ) = [10 * weight (kg)] + [6.25
* height (cm)] - [5 * age (y)] - 161
✗
-50 20 25 30 35 40
0
30
60
90
Pregnancy week
53 %
(+/- 2)
REEChange(%)
57 (+/-13) 40
(+/- 4)
Cold/Flu
After birth
41%
0 20 25 30 35 40
42
44
46
48
50
52
54
56
58
60
Weight(kg)
Pregnancy weeks
After birth
53
Baseline ~ 44 kg
Cold/Flu
REE does not follow the simple math of “higher
mass -> higher metabolic rate” from the
equation
Risk of
underfeeding
Risk of
over-
feeding
23
24. Comparison of REE with
Body Composition
Mifflin - St Jeor equation:
Woman:
REE(M-StJ) = [A * FFM (kg)] + [B * FM (kg)] + C
✗
0 20 25 30 35 40
10
20
30
40
50
60
11 kg
18 kg
53 kg
42 kg
39 kg
57 kg
Weight (kg)
Fat Mass (kg)
Lean Body Mass (kg)
Body(Total/Fat/Lean)
Mass(kg)
Pregnancy weeks
53 kg
44 kg
36.5 kg
7.5 kg
FFM
FM
-50 20 25 30 35 40
0
30
60
90
Pregnancy week
53 %
(+/- 2)
REEChange(%)
57 (+/-13) 40
(+/- 4)
Cold/Flu
After birth
41%
REE does not follow the simple math of:
“the higher the Free Fat Mass (FFM) or the
more Fat Mass (FM), the higher metabolic
rate” from an equation.
24www.breezing.co
25. Dr. St Jeor, creator of
MifflinSt Jeor’s REE
predictive equations
Picture of Dr. Sachiko St. Jeor at FNCE
2015, October 5th, using Breezing
Tracker
https://www.facebook.com/breezing.co
Dr. St Jeor is now a Breezing’s advocate
26. “ The use of predicative equations for
estimating REE are only ESTIMATIONS”
“We are much more complex as individuals and
the complexity is addressed only with a breath-
based REE measurement”
26www.breezing.co
27. What about weight management in sports?
= - [ + ]
Resting Activity
27www.breezing.co
28. Emily's goal:
• Needed to to reach 160 lbs
by competition day
•Bottom Line: Needed to lose
10 lbs in 2 months
Case #4 – Weight management in sports*
28
* Rich Wenner, athletes’ coach & Amber Yudell, nutritionist, Arizona State University
www.breezing.co
29. The results include all four module data from Breezing App
Resting Energy Expenditure (REE) (indirect calorimetry) Activity (manually entered), and assessed with HR monitor (PulseONE)
Diet (manually entered), and assessed with MyFitnessPal Weight (manually entered)
Case #4 – Weight management in sports
Weight (Lbs)
Resting Energy Expenditure (kcal/day)
0
500
1000
1500
2000
2500
Activity (kcal/day)
0
200
400
600
800
1000
1200
1400
1600
Calorie Intake (kcal/day)
0
500
1000
1500
2000
2500
154
156
158
160
162
164
166
168
170
172
Average: 1680 (sd: 130)
Average: 500 (sd: 290)
Average: 1720 (sd: 110)
TEE=REE+Act=2180 kcal/day
Intake= 1720 kcal/day
Deficit= -460 kcal/day
Competition day
-10lbs/9 weeks
29
30. http://instagify.com/media/980460235926117
550_1581604454
Emily J achieved her weight
goal of 160 lbs in 2 months,
and her life’s weightlifting
record (70 kg, 5Kg over
previous personal record)!
She can rescue someone
with her own weight now!
Case #4 – Weight management in sports
30www.breezing.co
32. Old REE measure initially
brought by the Breezing user
(2600 kcal/day)
Case #5 – Weight management in Hypothyroidism
-= [ + ]
✓ The user thought that he should be losing weight!
Case with Cytomel (Thyroid T3) - 25mcg/day
2050 kcal/day~ - 600 kcal/day
small kcal/day
32www.breezing.co
33. The new Breezing user got REE measurements for from Feb. 2nd to March 26th 2015 –
Total: 52 days
REE Mean
- 1SD
+1SD
Case #5 – Weight management in Hypothyroidism*
0 10 20 30 40
1000
1200
1400
1600
1800
2000
2200
2400
REE(kcal/day)
Days of testing (#)
REE Mean: 1730 kcal/day (SD: 200)
Relative Variability (68prob., =+/-1SD): +/- 11.5%
-= [ + ]
High variability was observed due
to the use of fast release of T3
hormone
Higher metabolic rate was
detected right after T3 hormone
intake
Despite the REE variability, an
average REE value could still be
defined
33
* Breezing’s user experience team. Advise from Dr. John Henried, MD, Sacramento, CA
35. Actual weight profile
241.3 241.3 241.9 241.9
220
225
230
235
240
245
250
255
260
23-03-2015 15:52 19-03-2015 09:51 17-03-2015 13:10 13-03-2015 06:56
Weight(lbs)
Day/Time
Weight profile showed less than 2% change, which corroborated the Energy Balance analysis
from Breezing
The REE average values adjusted the energy balance equation, despite the potential
hormonal variability.
Action: the user was switched to a slow release thyroid hormone to control the T3 levels in blood to
avoid spikes due to fast release
35www.breezing.co
36. • The breath measurement of Resting Energy Expenditure (REE) is
important to manage weight in a variety of different health-related
situations, including obesity, type 2 diabetes, hormonal problems,
pregnancy as well as in fitness training.
• The importance on breath analysis for REE is similar to a blood pressure
measurement for management of blood pressure.
• Calorie intake based on Resting Energy Expenditure measurement can
be accurately prescribed to manage weight successfully.
• Attempts to use an equation, instead of a measurement for Resting
Energy Expenditure, produce mere estimations (guess).
Conclusions from Case#1 – Case#5
36www.breezing.co
Blood pressure management Weight and wellness management
37. How we can increase metabolism and reverse
sedentary lifestyles without drastically altering our
schedules?
37
What about High Intensity Intermittent
Training (HIIT)?
38. Case #6: Study with High Intensity Intermittent
Training (HIIT)*
38
Total time = 4min
20s 20s 20s 20s 20s 20s 20s 20s
10s 10s 10s 10s 10s 10s 10s
Troy Anderson, Trainer
* In collaboration with Barb Ainsworth (Former ACSM President), Troy Anderson (CPT), and D. Jackemeyer (ASU)
39. Blood measurements
• Blood glucose
• Blood ketones
Metabolic measurements
• REE
• IEE pre exercise
• IEE post exercise
• IEE 1hr post exercise
• IEE 2hr post exercise
Body composition
• % Muscle mass
• % Fat mass
• %LBM
• Weight
• BMI
Intervention
19 subjects
Control
11 subjects
Intervention
24 subjects
Control
10 subjects
Total
enrolled
34
subjects
Random
allocation 4 subjects withdrew
1 moved to control
HIIT*
No Training
*3 HIIT sessions per week for 6 weeks
Case #6: Study Design*
HIIT
CONTROL
39
* In collaboration with Barb Ainsworth (Former ACSM President), Troy Anderson (CPT), and D. Jackemeyer (ASU)
40. Case #6: Quantification of the amount of exercise
Example: lifting work of 20 lbs
and 1.06 m with thruster
movement
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Power(W)
Work(J)&reps
Work and Power (Avgs) -- repeat ascending lifts
Erica - ef11 (per Kg Body Weight)
J x 10^-1 reps Watts
Session number
40* In collaboration with Barb Ainsworth (Former ACSM President), Troy Anderson (CPT), and D. Jackemeyer (ASU)
41. * Squatting work of 36-55 lbs
and 0.53 m with up & down
Resting
Pre exercise
0 hr post exercise
1 hr post exercise
2 hr post exercise
-1500
-1000
-500
0
500
1000
1500
2000
2500
3000
3500
Devon
REE(KCal/day)
REE:BL
REE:S1
REE:BL
REE:S1
*
REE/IEE(kCal/day)IEE(kCal/day)
HIIT day
HIIT day
No HIIT day
No HIIT day
41
Case #6: Quantifying Momentary Energy
Expenditure before and after exercising*
Can we detect a difference in metabolism between a High Intensity Interval
Training (HIIT) day vs a No-HIIT day ?
* In collaboration with Barb Ainsworth (Former ACSM President), Troy Anderson (CPT), and D. Jackemeyer (ASU)
42. Case #6: Effect of HIIT on individual’s energy expenditure
throughout training sessions
-350
-280
-210
-140
-70
0
70
140
210
280
350
AveragediEE(immpost)
**
*
HIIT
Control
No HIIT
A B C
*
D
AveragedIEEimmpost
Averaged change of pre- and post- energy expenditure (iEE = EEpost - EEpre) was
significantly different:
HIIT day vs. NO HIIT day (HIIT group)
HIIT day (HIIT group) vs. CONTROL (Control group)
42www.breezing.co
43. Is higher immediate post-exercise IEE change
related to muscle mass increase ?
43
?
IEEimm post
Muscle change (%)
44. The difference between groups
is significant at 80% level of
confidence
Case #6: Muscle Mass (%) Change & immediate post-exercise
Energy Expenditure Change (IEEimm post)
Group A: IEEimm post (HIIT with ≥6% muscle increase) = 241
kCal/day (SEM = 77)
Group B: IEEimm post (HIIT with (-1;4) % muscle increase) =
70 kCal/day (SEM = 58)
Difference A – B: 171 kCal/day
≥6% muscle
increase
=
44
46. 0 20 40 240 270
0
500
1000
1500
2000
2500
REE/IEE(kCal/day)
Day
HIIT HIIT
HIIT
HIITHIIT
HIIT
HIIT
Long-term RMR(REE) / IEE (MEE) tracking
Breezing personal parameter tracking of resting and High Intensity Interval
Training (HIIT) interventions: REE and IEE (MEE) values over nine months,
including seven HIIT session.
Case #7: Breezing Personal Tracking for over nine months
47. Fat Oxidation Dominant
HIIT
HIIT
0 200 400
0.65
0.70
0.75
0.80
0.85
0.90
0.95
RQ
Time (min)
Fat Oxidation Subordinant
One -day transient RQ during a fasting day
Case #7: Breezing Personal Tracking
Breezing personal parameter tracking of resting and High Intensity Interval
Training (HIIT) interventions: RQ values for 2 HIIT sessions over 6 hours in a
fasted individual.
48. HIIT
HIIT
Baseline: ~1,550 kCal/day
0 200 400
0
400
800
1200
1600
2000
2400
HH:MM
Time (min)
HH:MM
19:00
REE/IEE(kCal/day)
13:00
HIIT HIIT
One -day REE / IEE tracking
One -day cumulative EPOC
0 200 400
0
5
10
15
20
25
CumulativeEPOC(kCal)
Time (min)
Breezing personal parameter tracking
of resting and High Intensity Interval
Training (HIIT) interventions: REE and
IEE, and corresponding cumulative
EPOC parameters for 2 HIIT sessions
over 6 hours in a fasted individual.
Breezing Personal Tracking
48www.breezing.co
49. Summary of Case #6 & Case#7
Personalized tracking of metabolism (REE, RQ) in connection with
physical activity energy expenditure is possible
Metabolism change from our lifestyles changes can be quantified
Metabolism
(kCal/day)
Resting (RMR)
Time
Moment (MEE)
Fat
Carbs
Energy Source (RQ)
49
50. 0 30 60 300 330 360
0
500
1000
1500
2000
2500
3000
3500
Day of Intervetion (#)
REE(kCal/day)
0 30 60 300 330 360
0.6
0.7
0.8
0.9
1.0
RespiratoryQuotient
0.6
0.7
0.8
0.9
1.0
0 30 60 300 330 360
80
82
84
86
88
90
Weight(kg)
Day of intervention (#)
March 2014 to June 2015
(ketogenic diet- higher fat)
Jan. 2015 to April 2015
(ketogenic diet- lesser
fat.
Diet A: Ketogenic diet- higher fat:
Intake: 1800 cal/day,
Fat: 1250 cal (140g),
Protein: 360 cal (90g),
Carb: 180 cal (45g).
Diet B: Ketogenic diet- lesser fat:
Intake: 1200 – 1400 cal/day
Fat: 75 g,
Protein: 80g,
Carb: 5 days 50 g, 2 days 100g.
Diet A increased metabolic rate above 2,000 kcal/day
level, and Respiratory Quotient (RQ) reflected diet
composition.
Diet B did not change metabolic rate, it increased RQ
1, indicating only carbohydrate oxidation source.
Refs. for RQ values:
0.60 to 0.80: mostly fat oxidation
0.80 to 0.90: mixed source, fat and carb oxidation
0.90 to 1.00: mostly carbohydrate oxidation or
anaerobic metabolism increased.
Case #8: Long-Term
Resting Energy
Expenditure monitoring
on Ketogenic Diets
50
51. By knowing true REE and adding this information to
the user profile, we can make Activity Tracking
(calories burned from different activities) more
accurate.
51
www.breezing.co
In Conclusion
Editor's Notes
Group A: 11% of cases had measured REE values that were 200-500 kCal/day larger than the corresponding calculated REE values. This was indicative that a group of the participants had significant higher metabolic rates compared to a normal population of their same gender, age, weight and height.
Group B: 47% of cases had measured REE values within +/- 200 kCal/day compared to calculated REE values. This was indicative of metabolic rates within the expected values for equivalent population.
Group C: The remaining 42% of cases had measured REE values that were 200 - 1100 kCal/day smaller than the calculated REE values. These metabolic rates were unusually low compared to normal population of the same gender, age, height, and weight.
After Warming up of 5-10 min (60% max HR) :
20-seconds of work with 10-seconds rest performed for 8 rounds
Work to rest ratio of 2:1 (the work interval is 2 x the length of the rest interval)
Total time : 8 rounds = 4-minutes